TWIN-IT

The TWIN-IT project aims to address major bottlenecks in soft tissue injury prevention by developing personalized, multiscale digital twins that explicitly integrate the complex mechanobiological processes leading to pressure ulcers. Soft tissue damage arises from interactions across spatial and temporal scales: mechanical stresses disturb microvascular flow, reduce oxygen delivery, and trigger local inflammatory cascades, ultimately compromising tissue integrity. Current clinical practices rely on empirical thresholds and largely overlook these coupled phenomena, failing to account for patient-specific variability in tissue structure and physiological responses.

Our central hypothesis is that these mechanobiological pathways can be quantitatively modeled to predict the early onset of tissue damage and inform more effective prevention.

TWIN-IT builds on multiphase poromechanical frameworks grounded in thermodynamically constrained averaging (TCAT), extending them to capture oxygen transport and inflammation. A critical challenge lies in describing how sustained loading induces hypoxia and inflammatory signaling in a way that can be personalized and simplified for clinical integration.

Methodologically, TWIN-IT will establish a novel experimental platform combining controlled mechanical loading with high-resolution biochemical characterization (via LC-OCT and Raman spectroscopy) on human and reconstructed tissues. Parallel murine models will elucidate how prolonged loading disrupts microcirculation and triggers molecular changes, validated by histology and immunohistochemistry. These data will inform multiscale models coupling tissue deformation, interstitial fluid flow, and metabolic stress. Sensitivity analyses will isolate the most influential parameters, guiding the reduction into a low-fidelity digital twin that merges physics-based components with machine learning for real-time prediction.

This digital twin will be tested in transfemoral amputees, where prosthetic sockets impose complex load distributions. Embedded sensors will monitor pressure, temperature, and microcirculatory changes during gait, allowing assessment of model predictions and potential refinement of socket design. Additionally, TWIN-IT will explore soft interface materials to mitigate local peak loads, enhancing tissue protection based on digital twin insights.

Clinically, this approach addresses a profound need: prosthetic users often face discomfort, recurrent skin breakdown, and may abandon their devices, while existing prevention strategies rarely account for the underlying mechanobiology. By explicitly linking mechanical loads to hypoxia, inflammation, and early molecular changes, TWIN-IT aims to enable earlier, personalized intervention—reducing complications, improving device acceptance, and ultimately enhancing patient quality of life.

The consortium integrates ENSAM, Mines Paris PSL, TIMC, LBTI, CEA, clinical partners (Pôle Saint-Hélier, Fondation Hopale), and industrial collaborator PROTEOR, ensuring rigorous mechanistic modeling, advanced imaging, clinical validation, and translational impact. By addressing how mechanical loading evolves into biochemical and inflammatory dysfunction, TWIN-IT directly aligns with the PEPR Digital Health framework, targeting functional multiscale digital twins for early detection and prevention. This project aims for preclinical readiness by 2030, supporting both national and European priorities in personalized, data-driven healthcare.